Blog
How a premium card pays for its lounges, credits and perks
A model of 1,000 invented cardholders shows how the annual fee, interchange on big spending and unused credits fund the perks, and what that implies for you.
A premium card hands you a long list of benefits: airport lounges, annual credits, travel protections, concierge services, a high earning rate on certain spending. Then it charges a fee that can run into the hundreds of dollars. It is natural to wonder how the bank comes out ahead. The short answer is that it is pricing the card for a typical cardholder, and a typical cardholder does not collect everything on the list.
This post builds a simple model of how the numbers can work, using 1,000 invented cardholders and round figures, and then uses the model to explain what you see: why the fee is high, why credits come with conditions, why lounges can be crowded and why the fee tends to rise when benefits are added. The numbers are illustrations of the logic. They are not any issuer's actual economics, which are private, and they do not describe any specific card. For current fees and benefits, use the card's own profile, such as Chase Sapphire Reserve or The Platinum Card from American Express, and the issuer's published terms. Press coverage in 2025 reported that fees on some well-known premium cards were raised alongside new credits and benefits, which is one reason to check the current figures rather than trust anything you read from an earlier date.
What makes a card "premium"
A premium card is a rewards card, usually travel-oriented, that charges a high annual fee and provides a larger bundle of benefits than a mainstream card. The bundle typically includes some mix of:
- A higher earning rate in certain categories such as travel or dining.
- Annual credits, usually delivered as a statement credit, that offset specific purchases.
- lounge access through the issuer's own lounges or a network such as Priority Pass.
- Travel and purchase protections, which come from the card's guide to benefits and not from federal law.
- Status, such as a mid-tier hotel or airline tier.
- A metal card or other physical cues, which are a design choice and not a financial benefit.
Every item is paid for, either by the cardholder through the fee, by merchants through interchange or by the issuer out of its own margin. The next sections follow each source of money.
The first source: the fee
An annual fee is the simplest income. It is paid whether or not you use anything, it arrives at a known time, and it does not depend on how much you spend. If an issuer has 1,000 cardholders and the fee is $700, an invented figure, the fee alone brings in 1,000 × $700 = $700,000 a year.
The fee also screens. A cardholder who pays $700 a year has chosen to be in a group that expects to use the card heavily, and the group is the one the issuer wants: people who spend a lot, tend to pay in full and respond to offers. The fee works as a price for access to a set of benefits, and it carries information about the person who pays it.
The second source: interchange on large spending
Heavy spenders generate heavy interchange. Say the average cardholder in this model puts $60,000 a year on the card. If the card's interchange, the share of each purchase that goes to the issuer from the merchant's fee, averages an invented 2.4 percent, which is higher than a standard card because premium cards generally sit in higher interchange categories, then each cardholder generates $60,000 × 2.4% = $1,440. For 1,000 cardholders that is $1,440,000. The interchange fee entry and where cash back comes from, and why it is never really free money cover where that money comes from and why it is paid.
Interchange depends on spending, so it favors the cardholder who puts everything on the card. A cardholder who spends $15,000 a year generates a quarter of that amount, $360, and the model's issuer earns far less from them. This is why premium cards are marketed to heavy spenders and why a low spender can find the card poor value even with the benefits.
The third source: interest from the few who carry a balance
Premium cards tend to be held by people who pay in full, but not all do. Suppose 15 percent of cardholders carry a balance, averaging $5,000 over the year at 22.36 percent, which is the Federal Reserve's reported average APR on accounts assessed interest in the third quarter of 2026. Each of those pays $5,000 × 0.2236 = $1,118 in interest. For 150 cardholders that is 150 × $1,118 = $167,700.
Interest is a small share here, but it is large per person and it matters to the margin. It is also a reminder that the benefits are not cheaper for a cardholder who pays interest: the interest costs more than any perk is worth.
The costs the issuer carries
On the other side of the ledger, the issuer pays for the rewards, the credits, the lounges and the running of the card. The invented costs per cardholder:
- Rewards. The average cardholder earns 1.8 points per dollar on $60,000, which is 108,000 points. If the issuer's cost per point is an invented 1 cent, the cost is $1,080 per cardholder, or $1,080,000 across the group.
- Credits. The card advertises $500 a year in credits. Cardholders use 70 percent of them on average, an invented rate. Cost: $500 × 70% = $350 per cardholder, or $350,000.
- Lounges. Cardholders visit an average of four times a year, and the issuer pays an invented $30 per visit to the lounge operator. Cost: 4 × $30 = $120 per cardholder, or $120,000.
- Everything else. Fraud, servicing, credit losses and the cost of funding the card. An invented $400 per cardholder is $400,000.
The model in one table
| Line | Per cardholder | 1,000 cardholders |
|---|---|---|
| Annual fees | $700 | $700,000 |
| Interchange | $1,440 | $1,440,000 |
| Interest | $167.70 | $167,700 |
| Revenue | $2,307,700 | |
| Rewards | $1,080 | $1,080,000 |
| Credits used | $350 | $350,000 |
| Lounge visits | $120 | $120,000 |
| Other costs | $400 | $400,000 |
| Costs | $1,950,000 | |
| Margin | $357,700 |
The margin is $357,700, which is about 15.5 percent of revenue ($357,700 ÷ $2,307,700). The issuer in this model comes out ahead, and it does so even though the card gives each cardholder $1,080 of rewards, $350 of credits used and four lounge visits on top of the protections.
What moves the margin
The more useful exercise is to change one input at a time.
- Everyone uses every credit. Credits cost $500 each instead of $350. Costs rise by $150,000 and the margin falls to $207,700. Unused credits, which the industry calls breakage, are worth $150,000 to the issuer in this model.
- Nobody carries a balance. Interest revenue falls to zero and the margin falls to $190,000. A card that depends mostly on full-payers has to make up the gap elsewhere.
- Lounge visits double. Eight visits at $30 add $120,000 to costs and the margin falls to $237,700. Lounge use matters to the issuer's cost, which is a reason some issuers limit guests or require conditions for access. Read the card's guide to benefits for the current rules.
- The fee falls to $400. Revenue falls by $300,000 and the margin to $57,700. Without a high fee the model barely works, which is why generous benefits and low fees do not usually appear together.
- Spending drops to $30,000. Interchange halves to $720,000, rewards fall to $540,000, and the margin changes by (−$720,000 + $540,000) = −$180,000, falling to $177,700. A low-spending cardholder generates less revenue but also less reward cost.
Each of these is a lever the issuer can adjust. If credits are used too much, it can narrow them. If lounges are crowded, it can add conditions. If the margin is thin, it can raise the fee. That sequence is the logic behind many of the changes premium cardholders read about, and it is why a card you chose last year may not have the same fee, credits or access next year.
Three cardholders, one card
The issuer prices for the average. You live at one point on the spread. Here are three invented cardholders with the same card, using the model's $700 fee, and valuing points at an invented 1.25 cents each. Lounge visits are valued at $35, the invented price you would have paid for comparable access, and credits at the amount actually used, since an unused credit returns nothing.
| Annual spending | Points value | Credits used | Lounge visits | Total benefit | Net of $700 fee | Flat 2% cash back | Premium minus flat | |
|---|---|---|---|---|---|---|---|---|
| A, heavy spender | $80,000 | $1,800 | $450 | 6 ($210) | $2,460 | $1,760 | $1,600 | +$160 |
| B, moderate spender | $30,000 | $675 | $250 | 3 ($105) | $1,030 | $330 | $600 | −$270 |
| C, light spender | $12,000 | $270 | $100 | 1 ($35) | $405 | −$295 | $240 | −$535 |
How the first row is built: $80,000 × 1.8 points per dollar is 144,000 points, and 144,000 × $0.0125 = $1,800. The credits add $450 and six lounge visits add 6 × $35 = $210, which is $2,460 in all. After the $700 fee the net is $1,760, and a flat 2 percent card on the same $80,000 would pay $1,600, so the premium card is ahead by $160.
The pattern is the point. The card is ahead only for the cardholder who spends a great deal, uses most of the credits and visits lounges, and even then the margin over a plain cash back card is modest on these assumptions. For the moderate spender it is behind by $270 a year, and for the light spender by $535. Change the value of a point to 2 cents, by redeeming through transfers that you can really use, and Cardholder B's points are worth 54,000 × $0.02 = $1,080, which makes the total $1,080 + $250 + $105 = $1,435, a net of $735 and $135 ahead of the $600 from cash back. The value of a point, and how well the credits match your life, are the two inputs that swing the answer.
What the model leaves out
A model with only fees, interchange, interest and a handful of costs misses several things that matter to some cardholders and not to others:
- Protections. Travel delay and cancellation coverage, rental car coverage and purchase protection are benefits of the card, set out in its guide to benefits, and they can be worth a lot on the day you need them. Their terms, caps and exclusions vary, and the guide to benefits is the document to read. They are hard to price in advance because you do not know whether you will make a claim.
- Status and perks tied to a program. A hotel or airline tier has value only if you use it. what elite status from a hotel credit card does and does not get you shows the arithmetic for one kind of status.
- Convenience. Some cardholders value a concierge line, a smoother airport experience or the sense that a problem will be fixed quickly. Those are real preferences and hard to put into dollars.
- Behavior. A card with a $700 fee can make you spend more to feel you are getting value. That cost belongs in the tally.
The honest approach is to price what you can, name what you cannot and decide whether the unpriced part is worth the remaining gap.
Breakage: why credits come with conditions
Breakage is the part of a benefit that goes unused. A $15 monthly credit that you remember to use in nine months out of twelve is worth $135 to you and costs the issuer $135, not the $180 on the page, and the $45 difference is breakage. That is not a trick; it is how any benefit with a use-it-or-lose-it design works, and the design is in the open if you read the terms.
Several design choices increase breakage, and they are worth recognizing:
- Credits that reset monthly or quarterly, so that a missed period is lost, not carried forward.
- Credits that require enrollment, or must be used at a specific merchant, in a specific app or on a specific type of purchase.
- Credits that appear as a statement credit after the purchase, which means you have to notice that they posted.
- Credits tied to spending you would not otherwise do, where the cost to you is the purchase itself.
Whether this is fair is a judgment call. The credit is a real benefit for someone who would buy the item anyway. For someone who would not, it is a reason to spend that costs more than it returns, and the next post, premium card credits you will use versus credits you will chase, is a checklist for sorting the two.
Where the merchants fit
Some credits are linked to a particular merchant, retailer or travel brand. In some arrangements the partner may fund part of the benefit in return for customers, which makes the credit cheaper for the issuer than its face value suggests. The terms of these arrangements are private and differ widely, so treat that as a possibility and not as a fact about any given card. What it means for you is limited: the face value of a credit is the most it can be worth to you, and you collect it only by making the purchase.
What this means for you
The model is an argument for pricing a premium card against your own use, not against the list in the brochure. These are the questions it suggests:
- How much do I put on the card? Heavy spenders generate the revenue the model depends on, and they earn the most points back.
- Which credits would I use anyway? Count those at their face value and ignore the rest. how to count statement credits against an annual fee honestly shows the method.
- How often would I really use lounges, and at what price would I otherwise pay for the same comfort? is airport lounge access worth paying an annual fee for? covers the comparison.
- Do I carry a balance? If you do, interest erases the premium benefits.
- What would I do if the fee went up next year? Cards change, and the annual fee break-even calculator lets you test a new fee quickly.
Key takeaways
- A premium card is funded by three streams: the annual fee, interchange on heavy spending and, for some cardholders, interest. Rewards, credits, lounges and operating costs are paid from them.
- In the invented model of 1,000 cardholders the issuer earned a margin of $357,700, about 15.5 percent of revenue, even while paying each cardholder $1,080 of rewards.
- Unused credits are part of the pricing. In the model, breakage was worth $150,000 and the whole margin would be $207,700 without it.
- Levers such as the fee, credit design and lounge conditions explain why premium cards change over time, and why a high fee and generous benefits go together.
- Price the card against your own spending and use, count only credits you would use anyway and do not carry a balance on it.